Skip to Main Content
PreviousNext

Peer-graded Assignment: Programming task

i
 
It looks like this is your first peer-graded assignment.  Learn more
  1. Instructions
  2. My submission
  3. Discussions
Final Programming Assignment - Calculus and Optimization for Machine Learning (Gradient Descent Implementation)
Submitted on April 29, 2020
Shareable Link
Prompt

Week 6. Optimization. Programming Task.

Final Programming Assignment - Calculus and Optimization for Machine Learning (Gradient Descent Implementation)
Final Programming Assignment - Calculus and Optimization for Machine Learning (Gradient Descent Implementation)
This assignment contains, 1. Reading and Preparing 2. A function to compute the Loss(w), given the weights, features and actual house prices 3. Analytical computation the gradient of the Loss(w) 4. A function to compute the gradient of the Loss function at a given point 5. Implementation of gradient descent 6. Answers to discussion questions
Rubric

Task 1. Reading and Preparing.

Does checker say 'All passed'?

0 points

No

1 point

Yes

Task 2. Compute analytically the Loss(w⃗)Loss(\vec{w}) function.

Does checker say 'All passed'?

0 points

No

1 point

Yes

Task 3. Compute analytically the gradient of the Loss(w⃗)Loss(\vec{w}).

Is the gradient correct?

0 points

No

1 point

Yes

Task 4. Write a function to compute the gradient of the Loss function in the given point.

Does checker say 'All passed'?

0 points

No

1 point

Yes

Task 5. Write gradient descent.

Check for at least 2 experiments. Check if at least one graph converges.

0 points

Out of all experiments zero converged. Zero graphs provided.

0.5 points

Only one experiment is done.

1 point

At least 2 experiments are done. At least one graph converges.

Task 6. Discussion.

0 points

Zero questions answered.

0.3 points

Question #1 is answered.

0.4 points

Only one of the questions #2, #3 or #4 is answered.

0.7 points

Question #1 is answered and only one of the questions #2, #3 or #4 is answered.

0.8 points

Two of the questions #2, #3 or #4 are answered.

1.1 points

Question #1 is answered and Two of the questions #2, #3 or #4 are answered.

1.2 points

All of the questions #2, #3 or #4 are answered.

1.5 points

All of the questions are answered.

How could your peer improve this submission even further?

Stamatis Mastromichalakis
Its excellent work
Andrey Vershinin
Seems you forgot to normalize X in Task 5, that's why predicted prices are in 1e3 range, while actual in 10s

Comments

Comments left for the learner are visible only to that learner and the person who left the comment.

Confirm Navigation

Are you sure you want to leave this page?
  

Confirm Navigation

Are you sure you want to leave this page?